[vlc-devel] [RFC/WIP PATCH 0/3] decoder-packetizer chicken-egg situation

[Thomas Guillem]

On Fri, Jul 1, 2016, at 14:41, Rémi Denis-Courmont wrote:
> Le 2016-07-01 12:18, Thomas Guillem a écrit :
> > This set of patches is a (work in progress) attempt to fix the 
> > chicken-egg
> > situation of decoder-packetizer.
> 
> I don't understand what problem this solves. Furthermore, I think 
> adding a new decoder type/capability is a very bad idea because modules 
> of different capabilities cannot be sorted/probed by priority.

Ah yes, this can't work, I didn't think about capabilities.

The problem I wanted to solve: having decoder modules that depend on a
specific packetizer. This could simplify dts/a52 packetizers/decoder
mess.

With some format, both the A52 packetizer and the A52 decoder are used
and are doing exactly the same thing. One solution is to have the a52
decoder depending on the A52 packetizer and don't re-packetize in the
decoder callbacks.

> 
> -- 
> Rémi Denis-Courmont
> http://www.remlab.net/
> _______________________________________________
> vlc-devel mailing list
> To unsubscribe or modify your subscription options:
> https://mailman.videolan.org/listinfo/vlc-devel